The ‘microfoundations’ metaphor, used as an explanatory strategy in mainstream macroeconomics, seeks to reduce macroeconomics to statements about the aggregate behaviour of individuals. Viewed as representative agents with rational expectations, their quest to maximise lifetime utility, subject to shocks in a Dynamic Stochastic General Equilibrium (DSGE) framework, is seen as the basis of macroeconomic phenomena.

In Reference KingKing (2012), I set out reasons why this explanatory approach is doomed to failure, and argue that heterodox economists of all persuasions, almost none of whom support the strategy, should desist from using the metaphor. Several pages in the book are devoted to a dissection of the use of the microfoundations metaphor by Marxists, the most prominent being John Roemer and Jon Elster (Reference KingKing, 2012: 141–145). It seems that Reference RoemerRoemer (1979) was the first to use the word ‘microfoundations’ in the title of an article, and in his influential book, Analytical Foundations of Marxian Economic Theory, he advocated ‘deriving the aggregate behavior of the economy as a consequence of the actions of individuals’ (Reference RoemerRoemer, 1981: 7). Elster went even further in his defence of a micro-reductionist approach to Marxian social theory. By methodological individualism, he wrote,

This was the same Elster who in an earlier work had pointed to the importance of avoiding fallacies of composition in social theory! The ‘Rational Choice Marxism’ that was advocated by Roemer and Elster did not win many converts, and it has largely been abandoned even by those who continue to describe themselves as ‘analytical Marxists’ (Reference VenezianiVeneziani, 2012). But the term ‘microfoundations’ is still used by some Marxists, and micro-reduction continues to attract some support from radical political economists: see, for example, the chapters by Jonathan Goldstein and Bill Gibson in Reference Goldstein and HillardGoldstein and Hillard (2009).

There are several reasons why the micro-reduction project in economics cannot work. Each of them appears sufficient on its own to rule out micro-reduction. One, the existence of emergent properties rests on the complexity of real-world economies, as Bernd-Olaf Küppers explains,

(Küppers’ use of the word ‘determines’ is probably too strong here, ‘significantly affects’ would suffice). There is growing recognition of the role of emergent properties in economics, for example, in the relationship between investment and productivity (Reference Harper and EndresHarper and Endres, 2012).

Now Küppers’ ‘two theses’ also apply to much simpler systems, in which complexity is not an issue. The first, the fallacy of composition, is not discussed here. I focus instead on the second thesis, which provides another reason why micro-reduction is doomed to failure: downward causation. It is very simple to summarise, and very difficult to deny. We, as individuals, are very heavily influenced by society. Our tastes and preferences, our values and beliefs, our knowledge and productive skills, depend on the society that we live in and the economy that is embedded in it. But foundations are not like this. They are put in place first, and then the building is constructed on top of them. In the process of construction, the foundations do not change. After the construction is completed, the building may be altered in various ways (redecorated, refurbished, renovated, reconstructed to a great or lesser extent) without occasioning any alterations to the foundations. That is what foundations are. But humans are not like that, and the microfoundations metaphor is therefore misconceived.

In the other social sciences, downward causation is an accepted fact, almost a commonplace. I will confine myself to just one example, which I came upon by chance, late in 2014 when I was writing this article. The Belgian psychologist, Paul Verhaeghe, has documented the consequences for the human personality of the rise of neoliberalism in his book (co-authored with Jane Hedley-Prole), What about Me? The Struggle for Identity in a Market-Based Society, and has summarised the arguments in a brief but incisive article in the Guardian Weekly (Reference VerhaegheVerhaeghe, 2014; Reference Verhaeghe and Hedley-ProleVerhaeghe and Hedley-Prole, 2014;). He shows how a competitive economic system that rewards psychopathic personality traits has changed our ethics and our personalities. In a neoliberal meritocracy, solidarity becomes an expensive luxury, and has given way to temporary alliances used to extract profit. Social ties with colleagues are weakened, workplace bullying becomes widespread and people’s self-respect is seriously damaged. Social norms and values are being changed, Verhaeghe concludes, and not for the better. On the contrary, the new economic system is bringing out the worst in us.

Downward causation means that micro-reduction is most unlikely to succeed in the social sciences (least of all in economics). Similar explanatory projects in the natural sciences have not been very successful (Reference KingKing, 2012, chapters 3–4). The philosopher, Steven Horst, lists some of the frequently discussed ‘cases of failed or incomplete intertheoretic reduction’ in the scientific literature. They include the reduction of thermodynamics to statistical mechanics, the reduction of thermodynamics/statistical mechanics to quantum mechanics, the reduction of chemistry to quantum mechanics, the reduction of classical mechanics to quantum mechanics, the reduction of evolutionary biology to genetics, and the reduction of genetics to molecular genetics (Reference HorstHorst, 2007: 59).

The case of genetic reduction is especially interesting, since it has been fatally undermined by downward causation. Critics of Richard Dawkins’ ‘hierarchical reductionism’ argue that what genes do depends on the organisms in which the genes are located and the natural and social environment in which these organisms live. Thus, Reference Gould, McGarr and RoseStephen Jay Gould (2007) rejects Dawkins’ argument that

The physiologist, Denis Noble, uses another striking metaphor. In a discussion with journalist Reference CraceJohn Crace (2008), he ‘reverses Dawkins’ ‘selfish gene’ image (‘They are in you and me; they created us, body and mind; and their preservation is the ultimate rationale for our existence’). Noble writes instead of the ‘prisoner gene’:

There are fewer advocates of genetic reductionism today than was once the case, owing to a growing recognition of the importance of downward causation – the influence of organisms and environment on the behaviour of genes.

Returning to economics, Peter Kriesler argues convincingly that Michał Kalecki was right on the question of microfoundations, as on several other important issues of economic theory and methodology. Kalecki maintained that macroeconomics and microeconomics should be thought of as being in a horizontal rather than a vertical relationship, closely connected to and influencing each other but also relatively autonomous, with neither constituting the foundations of the other. For Kalecki,

In this respect, Kriesler concludes, Kalecki’s method is very similar to that of Ricardo and Marx (Reference Kriesler and KingKriesler, 1996: 67).

Some economists, usually but not always of a heterodox inclination, have responded to these issues by suggesting that we ought instead to be providing macrofoundations for microeconomics (Reference KingKing, 2012: 42–45). They include authors as diverse as Sergio Rossi, Jan Kregel, David Colander, Geoff Hodgson, Giuseppe Fontana and Julie Nelson (cited in Reference KingKing, 2012: 160–170). While I sympathise with this reaction to the significance of downward causation in economic life, I cannot agree with the ‘macrofoundations’ metaphor, which conjures up the bizarre image of a house with the foundations perched on top of the roof. It would be better to avoid foundational metaphors of any sort, I think. If constructional metaphors are unavoidable, it would be sensible to follow Kalecki and use them in the horizontal, not the vertical plane.

Reference PilkingtonPilkington (2013) agrees, but argues that the analysis needs to be taken further. In the early parts of my 2012 book, he argues, I suggest that we should place microeconomics and macroeconomics on an equal footing. But this is wrong, because ‘the group has more power over the individual than the individual has over the group’. This can be seen in the ‘paradox of thrift’, where

Again, I have some sympathy with this position. But I am concerned that the road that leads from individualism to holism might at some point become a slippery slope, propelling the unwary traveller into an extreme form of holism that is very difficult to justify.

Here, for example, is Reference Von MisesLudwig Von Mises (1949 [1996]), denouncing the notion that ‘society is an entity living its own life, independent of and separate from the lives of the various individuals, acting on its own behalf and aiming at its own ends which are different from the ends sought be the individuals’ (p. 145). I doubt whether many 21st century Marxists would adhere to this version of holism. It comes, admittedly, from the pen of a man who was not himself given to excessive moderation, and who is supposed to have walked out of a meeting of the Mont Pélerin Society muttering, ‘You’re all socialists!’. But it is not difficult to imagine, up in Moscow, in the spring of 1949, Stalinist ideologues denouncing bourgeois individualism in rather similar terms to those invoked by Mises. The temptation should certainly be avoided today.

All this brings us back to another very big question: the relationship between freedom and historical/social determinism in the thought of Karl Marx. It brings to mind the three-stage typology of human history that he set out in the Grundrisse:

It is obvious that there is very little freedom in the first stage, which refers to the slave and feudal modes of production: 98% of all individuals are the property of slave-owners, or tied to the land of their lords, and forced to perform unpaid surplus labour for them. Their lives are very comprehensively determined by the social relations of production.

Things seem different in the second stage, of ‘objective dependence’, in which commodity production (in particular, capitalist commodity production) has removed the blatant unfreedom of slavery and feudalism. But appearances are deceptive. As Marx put it, sardonically, the worker in a capitalist society is free in a double sense: not owned by anyone, but also not owning anything other than his or her capacity to work. ‘Free’ from the burdens of ownership, the workers are entirely dependent on the capitalists for the means of subsistence. Even the capitalists, however, only appear to be free. In fact, the pervasive phenomena of alienation and fetishism mean that the capitalist, just like the worker, is unfree. It could be argued that capitalists are actually less free than workers, who do at least realise (or used to realise) that something is badly wrong. The capitalist, too, is dominated by his own products, whether they be physical objects, or social relations, or false ideas about these social relations. Only in the communist future, when commodity production has been superseded by cooperation, alienation and fetishism have been abolished and objective or material dependence has been transcended – only then, in the third and final stage of human history, the stage of free individuality, can people really be free.

All this is well-known, and is often recounted in critical studies of Marx’s thought (see, for example, Reference Howard and KingHoward and King, 1985). But its implications were not always fully appreciated, especially during the dark days of Stalinist tyranny. One Marxist philosopher who had some very interesting things to say about capitalism and freedom was Herbert Marcuse. His argument in the classic Reference Marcuse1964 text One-Dimensional Man is lucidly summarised by Reference Kellner and MarcuseDouglas Kellner (1991 [2002]):

This is downward causation with a vengeance. I do not imagine that Marcuse ever commented on the microfoundations metaphor, which was only beginning to be widely used when he died, in 1979. I suspect that he would have given it very short shrift, along with the micro-reduction project that lay behind it.

I also do not know if Marcuse ever met Mises, but I can imagine what he would have had to say if he had ever found himself in the same room as the Austrian liberal philosopher. After all, Marcuse was ‘a radical individualist’ who was ‘deeply disturbed by the decline of the traits of authentic individuality that he so highly value[d]’ (Reference Kellner and MarcuseKellner, 1991 [2002]: xxviii). We share the same deep belief in the virtues of human freedom, he would have insisted to Mises. It is just that you seem to believe that it can be achieved under capitalism. I am telling you very clearly that it cannot, even for the capitalists, let alone for the workers. Real freedom requires the abolition of commodity production. We shall not see it, this side of the revolution.

And so, to conclude, I agree with Marcuse that micro-reduction is impossible under capitalism, and I am confident that he would have agreed with me that Marxists should therefore at present avoid the microfoundations metaphor. Note that this conclusion has been reached on the basis of downward causation alone, without fuller analysis of fallacies of composition or emergent properties latent in the complexities of social life. Micro-reduction may simply prove to be impossible in the social sciences, whatever the mode of production, so that the microfoundations metaphor might turn out to be unusable in all stages of human history. It would not be a very great loss.